Search

Your search keyword '"Kevin Dalleau"' showing total 10 results
Start Over Author "Kevin Dalleau" Search Limiters Full Text
10 results on '"Kevin Dalleau"'

1. Head-to-head comparison of clustering methods for heterogeneous data: a simulation-driven benchmark

Author

Gregoire Preud’homme, Kevin Duarte, Kevin Dalleau, Claire Lacomblez, Emmanuel Bresso, Malika Smaïl-Tabbone, Miguel Couceiro, Marie-Dominique Devignes, Masatake Kobayashi, Olivier Huttin, João Pedro Ferreira, Faiez Zannad, Patrick Rossignol, and Nicolas Girerd

Subjects
Medicine, Science
Abstract

Abstract The choice of the most appropriate unsupervised machine-learning method for “heterogeneous” or “mixed” data, i.e. with both continuous and categorical variables, can be challenging. Our aim was to examine the performance of various clustering strategies for mixed data using both simulated and real-life data. We conducted a benchmark analysis of “ready-to-use” tools in R comparing 4 model-based (Kamila algorithm, Latent Class Analysis, Latent Class Model [LCM] and Clustering by Mixture Modeling) and 5 distance/dissimilarity-based (Gower distance or Unsupervised Extra Trees dissimilarity followed by hierarchical clustering or Partitioning Around Medoids, K-prototypes) clustering methods. Clustering performances were assessed by Adjusted Rand Index (ARI) on 1000 generated virtual populations consisting of mixed variables using 7 scenarios with varying population sizes, number of clusters, number of continuous and categorical variables, proportions of relevant (non-noisy) variables and degree of variable relevance (low, mild, high). Clustering methods were then applied on the EPHESUS randomized clinical trial data (a heart failure trial evaluating the effect of eplerenone) allowing to illustrate the differences between different clustering techniques. The simulations revealed the dominance of K-prototypes, Kamila and LCM models over all other methods. Overall, methods using dissimilarity matrices in classical algorithms such as Partitioning Around Medoids and Hierarchical Clustering had a lower ARI compared to model-based methods in all scenarios. When applying clustering methods to a real-life clinical dataset, LCM showed promising results with regard to differences in (1) clinical profiles across clusters, (2) prognostic performance (highest C-index) and (3) identification of patient subgroups with substantial treatment benefit. The present findings suggest key differences in clustering performance between the tested algorithms (limited to tools readily available in R). In most of the tested scenarios, model-based methods (in particular the Kamila and LCM packages) and K-prototypes typically performed best in the setting of heterogeneous data.

Published
2021
Full Text
View/download PDF

2. Learning from biomedical linked data to suggest valid pharmacogenes

Author

Kevin Dalleau, Yassine Marzougui, Sébastien Da Silva, Patrice Ringot, Ndeye Coumba Ndiaye, and Adrien Coulet

Subjects
Linked data, Pharmacogenomics, Data mining, Knowledge discovery from databases, Machine learning, Valid pharmacogenes, Computer applications to medicine. Medical informatics, R858-859.7
Abstract

Abstract Background A standard task in pharmacogenomics research is identifying genes that may be involved in drug response variability, i.e., pharmacogenes. Because genomic experiments tended to generate many false positives, computational approaches based on the use of background knowledge have been proposed. Until now, only molecular networks or the biomedical literature were used, whereas many other resources are available. Method We propose here to consume a diverse and larger set of resources using linked data related either to genes, drugs or diseases. One of the advantages of linked data is that they are built on a standard framework that facilitates the joint use of various sources, and thus facilitates considering features of various origins. We propose a selection and linkage of data sources relevant to pharmacogenomics, including for example DisGeNET and Clinvar. We use machine learning to identify and prioritize pharmacogenes that are the most probably valid, considering the selected linked data. This identification relies on the classification of gene–drug pairs as either pharmacogenomically associated or not and was experimented with two machine learning methods –random forest and graph kernel–, which results are compared in this article. Results We assembled a set of linked data relative to pharmacogenomics, of 2,610,793 triples, coming from six distinct resources. Learning from these data, random forest enables identifying valid pharmacogenes with a F-measure of 0.73, on a 10 folds cross-validation, whereas graph kernel achieves a F-measure of 0.81. A list of top candidates proposed by both approaches is provided and their obtention is discussed.

Published
2017
Full Text
View/download PDF

3. Machine Learning-Derived Echocardiographic Phenotypes Predict Heart Failure Incidence in Asymptomatic Individuals

Author

Kevin Dalleau, Jean-Marc Boivin, Malika Smaïl-Tabbone, Anne-Cecile Huby, Patrick Rossignol, Margret Leosdottir, Marie-Dominique Devignes, Gregoire Preud'homme, Olivier Huttin, Emmanuel Bresso, Martin Magnusson, Faiez Zannad, Kevin Duarte, João Pedro Ferreira, Masatake Kobayashi, Stanislas study, Erwan Bozec, Peter M. Nilsson, Zohra Lamiral, Nicolas Girerd, Défaillance Cardiovasculaire Aiguë et Chronique (DCAC), Centre Hospitalier Régional Universitaire de Nancy (CHRU Nancy)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Université de Lorraine (UL), Centre d'investigation clinique plurithématique Pierre Drouin [Nancy] (CIC-P), Centre d'investigation clinique [Nancy] (CIC), Centre Hospitalier Régional Universitaire de Nancy (CHRU Nancy)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Université de Lorraine (UL)-Centre Hospitalier Régional Universitaire de Nancy (CHRU Nancy)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Université de Lorraine (UL), Cardiovascular and Renal Clinical Trialists [Vandoeuvre-les-Nancy] (INI-CRCT), Institut Lorrain du Coeur et des Vaisseaux Louis Mathieu [Nancy], French-Clinical Research Infrastructure Network - F-CRIN [Paris] (Cardiovascular & Renal Clinical Trialists - CRCT ), Lund University [Lund], Skane University Hospital [Malmo], Computational Algorithms for Protein Structures and Interactions (CAPSID), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Complex Systems, Artificial Intelligence & Robotics (LORIA - AIS), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), The STANISLAS Cohort visit 4 was sponsored by the Nancy CHRU and was funded in part by the Programme Hospitalier de Recherche Clinique Interrégional. Biomarker studies are co-funded by the French National Research Agency Fighting Heart Failure (ANR-15-RHU-0004) and FEDER Lorraine, and all French co-authors are supported by the French Programme d'investissements d'avenir project 'Lorraine Université d’Excellence' GEENAGE (ANR-15-IDEX-04-LUE) programs, and the Contrat de Plan Etat Région Lorraine and FEDER IT2MP. The research leading to these results also received support from the European Union Commission’s Seventh Framework program under grant 305507 (Heart OMics in Aging). Support was also provided from the 'EXPERT' ERA-CVD 2016 and MR-Focus (both grants managed by the French National Research Agency). Drs. Girerd, Rossignol, and Zannad are supported by the French National Research Agency Fighting Heart Failure (ANR-15-RHU-0004), the French PIA project Lorraine Université d’Excellence GEENAGE (ANR-15-IDEX-04-LUE) programs, and the Contrat de Plan Etat Région Lorraine and FEDER IT2MP., ANR-15-RHUS-0004,FIGHT-HF,Combattre l'insuffisance cardiaque(2015), ANR-15-IDEX-0004,LUE,Isite LUE(2015), European Project, European Project: 305507,EC:FP7:HEALTH,FP7-HEALTH-2012-INNOVATION-1,HOMAGE(2013), European Project: ERA-CVD/0001/2016,MINOTAUR, BOZEC, Erwan, Combattre l'insuffisance cardiaque - - FIGHT-HF2015 - ANR-15-RHUS-0004 - RHUS - VALID, ISITE - Isite LUE - - LUE2015 - ANR-15-IDEX-0004 - IDEX - VALID, Entreprises lorraines, parties prenantes et Développement Durable (FEDER) (Région Lorraine) - INCOMING, Heart OMics in AGEing - HOMAGE - - EC:FP7:HEALTH2013-02-01 - 2019-01-31 - 305507 - VALID, Metabolic Therapy for Managing Diastolic Heart Failure - MINOTAUR - ERA-CVD/0001/2016 - INCOMING, and Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)

Subjects
Male, medicine.medical_specialty, heart failure, 030204 cardiovascular system & hematology, Asymptomatic, Ventricular Function, Left, 03 medical and health sciences, 0302 clinical medicine, [SDV.MHEP.CSC]Life Sciences [q-bio]/Human health and pathology/Cardiology and cardiovascular system, Predictive Value of Tests, Internal medicine, medicine, Humans, Radiology, Nuclear Medicine and imaging, 030212 general & internal medicine, Aged, business.industry, Incidence (epidemiology), Incidence, biomarkers, Stroke Volume, Middle Aged, medicine.disease, Phenotype, echocardiogram, 3. Good health, cardiovascular diseases, [SDV.MHEP.CSC] Life Sciences [q-bio]/Human health and pathology/Cardiology and cardiovascular system, machine learning, Echocardiography, Heart failure, Female, prognosis, medicine.symptom, Cardiology and Cardiovascular Medicine, business, cluster analysis
Abstract

This study sought to identify homogenous echocardiographic phenotypes in community-based cohorts and assess their association with outcomes.Asymptomatic cardiac dysfunction leads to a high risk of long-term cardiovascular morbidity and mortality; however, better echocardiographic classification of asymptomatic individuals remains a challenge.Echocardiographic phenotypes were identified using K-means clustering in the first generation of the STANISLAS (Yearly non-invasive follow-up of Health status of Lorraine insured inhabitants) cohort (N = 827; mean age: 60 ± 5 years; men: 48%), and their associations with vascular function and circulating biomarkers were also assessed. These phenotypes were externally validated in the Malmö Preventive Project cohort (N = 1,394; mean age: 67 ± 6 years; men: 70%), and their associations with the composite of cardiovascular mortality (CVM) or heart failure hospitalization (HFH) were assessed as well.Three echocardiographic phenotypes were identified as "mostly normal (MN)" (n = 334), "diastolic changes (D)" (n = 323), and "diastolic changes with structural remodeling (D/S)" (n = 170). The D and D/S phenotypes had similar ages, body mass indices, cardiovascular risk factors, vascular impairments, and diastolic function changes. The D phenotype consisted mainly of women and featured increased levels of inflammatory biomarkers, whereas the D/S phenotype, consisted predominantly of men, displayed the highest values of left ventricular mass, volume, and remodeling biomarkers. The phenotypes were predicted based on a simple algorithm including e', left ventricular mass and volume (e'VM algorithm). In the Malmö cohort, subgroups derived from e'VM algorithm were significantly associated with a higher risk of CVM and HFH (adjusted HR in the D phenotype = 1.87; 95% CI: 1.04 to 3.37; adjusted HR in the D/S phenotype = 3.02; 95% CI: 1.71 to 5.34).Among asymptomatic, middle-aged individuals, echocardiographic data-driven classification based on the simple e'VM algorithm identified profiles with different long-term HF risk. (4th Visit at 17 Years of Cohort STANISLAS-Stanislas Ancillary Study ESCIF [STANISLASV4]; NCT01391442).

Published
2021
Full Text
View/download PDF

4. Unsupervised Extra Trees: a stochastic approach to compute similarities in heterogeneous data

Author

Kevin Dalleau, Miguel Couceiro, Malika Smaïl-Tabbone, Computational Algorithms for Protein Structures and Interactions (CAPSID), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Complex Systems, Artificial Intelligence & Robotics (LORIA - AIS), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL), Knowledge representation, reasonning (ORPAILLEUR), Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Kevin Dalleau’s PhD is funded by the RHU FIGHTHF (ANR-15-RHUS-0004) and the Region Grand Est (France)., ANR-15-RHUS-0004,FIGHT-HF,Combattre l'insuffisance cardiaque(2015), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), and Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)

Subjects
0301 basic medicine, Hierarchical agglomerative clustering, Unsupervised classification, Computer science, Decision tree, Monotonic function, Similarity measure, Clustering, 03 medical and health sciences, 0302 clinical medicine, [INFO.INFO-LG]Computer Science [cs]/Machine Learning [cs.LG], Robustness (computer science), Cluster (physics), Preprocessor, Cluster analysis, business.industry, Applied Mathematics, Pattern recognition, Computer Science Applications, 030104 developmental biology, Computational Theory and Mathematics, 030220 oncology & carcinogenesis, Modeling and Simulation, Artificial intelligence, Extremely randomized trees, business, Information Systems
Abstract

International audience; In this paper we present a method to compute similarities on unlabeled data, based on extremely randomized trees. The main idea of our method, Unsu-pervised Extremely Randomized Trees (UET) is to randomly split the data in an iterative fashion until a stopping criterion is met, and to compute a similarity based on the co-occurrence of samples in the leaves of each generated tree. Using a tree-based approach to compute similarities is interesting, as the inherent We evaluate our method on synthetic and real-world datasets by comparing the mean similarities between samples with the same label and the mean similarities between samples with different labels. These metrics are similar to intracluster and intercluster similarities, and are used to assess the computed similarities instead of a clustering algorithm's results. Our empirical study shows that the method effectively gives distinct similarity values between samples belonging to different clusters, and gives indiscernible values when there is no cluster structure. We also assess some interesting properties such as in-variance under monotone transformations of variables and robustness to correlated variables and noise. Finally , we performed hierarchical agglomerative clustering on synthetic and real-world homogeneous and heterogeneous datasets using UET versus standard similarity measures. Our experiments show that the algorithm outperforms existing methods in some cases, and can reduce the amount of preprocessing needed with many real-world datasets.

Published
2020
Full Text
View/download PDF

5. PGxCorpus, a manually annotated corpus for pharmacogenomics

Author

Kevin Dalleau, Joël Legrand, Adrien Coulet, Nadine Petitpain, Malika Smaïl-Tabbone, Romain Gogdemir, Yannick Toussaint, William Digan, Cédric Bousquet, Marie-Dominique Devignes, Patrice Ringot, Ndeye-Coumba Ndiaye, Chia-Ju Lee, Natural Language Processing : representations, inference and semantics (SYNALP), Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), Knowledge representation, reasonning (ORPAILLEUR), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), Laboratoire d'Informatique Médicale et Ingénierie des Connaissances en e-Santé (LIMICS), Institut National de la Santé et de la Recherche Médicale (INSERM)-Sorbonne Université (SU)-Université Sorbonne Paris Nord, Computational Algorithms for Protein Structures and Interactions (CAPSID), Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Complex Systems, Artificial Intelligence & Robotics (LORIA - AIS), Hôpital Européen Georges Pompidou [APHP] (HEGP), Assistance publique - Hôpitaux de Paris (AP-HP) (AP-HP)-Hôpitaux Universitaires Paris Ouest - Hôpitaux Universitaires Île de France Ouest (HUPO), Department of Biomedical Informatics and Medical Education, University of Washington, University of Washington [Seattle], Nutrition-Génétique et Exposition aux Risques Environnementaux (NGERE), Institut National de la Santé et de la Recherche Médicale (INSERM)-Université de Lorraine (UL), Centre Régional de PharmacoVigilance de Lorraine (CRPV Lorraine), Centre Hospitalier Régional Universitaire de Nancy (CHRU Nancy), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), ANR-15-CE23-0028, Agence Nationale de la Recherche, 15-IDEX-0004, Université de Lorraine, Snowball Inria Associate Team, GRID5000, ANR-15-CE23-0028,PractiKPharma,Confrontation entre connaissances de l'état de l'art et connaissances extraites de dossiers patients en pharmacogénomique(2015), Coulet, Adrien, Interactions humain-machine, objets connectés, contenus numériques, données massives et connaissance - Confrontation entre connaissances de l'état de l'art et connaissances extraites de dossiers patients en pharmacogénomique - - PractiKPharma2015 - ANR-15-CE23-0028 - AAPG2015 - VALID, CRHU Nancy, and Service Informatique de Soutien à la Recherche (SISR)

Subjects
Data Descriptor, Computer science, [INFO.INFO-TT] Computer Science [cs]/Document and Text Processing, computer.software_genre, [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI], 0302 clinical medicine, Resource (project management), Drug response, 030212 general & internal medicine, lcsh:Science, Data Curation, ComputingMilieux_MISCELLANEOUS, [INFO.INFO-BI] Computer Science [cs]/Bioinformatics [q-bio.QM], 0303 health sciences, 3. Good health, Computer Science Applications, [INFO.INFO-TT]Computer Science [cs]/Document and Text Processing, [SDV.SP.PHARMA] Life Sciences [q-bio]/Pharmaceutical sciences/Pharmacology, Supervised Machine Learning, Statistics, Probability and Uncertainty, Natural language processing, Information Systems, [INFO.INFO-AI] Computer Science [cs]/Artificial Intelligence [cs.AI], Statistics and Probability, PubMed, [SDV.SP.MED] Life Sciences [q-bio]/Pharmaceutical sciences/Medication, [SDV.GEN.GH] Life Sciences [q-bio]/Genetics/Human genetics, Library and Information Sciences, Domain (software engineering), Education, 03 medical and health sciences, [SDV.SP.MED]Life Sciences [q-bio]/Pharmaceutical sciences/Medication, Component (UML), Genetics, Humans, 030304 developmental biology, business.industry, Health care, Significant part, Precision medicine, ComputingMethodologies_PATTERNRECOGNITION, [SDV.GEN.GH]Life Sciences [q-bio]/Genetics/Human genetics, Pharmacogenetics, Pharmacogenomics, [SDV.SP.PHARMA]Life Sciences [q-bio]/Pharmaceutical sciences/Pharmacology, lcsh:Q, Artificial intelligence, [INFO.INFO-BI]Computer Science [cs]/Bioinformatics [q-bio.QM], business, computer
Abstract

Pharmacogenomics (PGx) studies how individual gene variations impact drug response phenotypes, which makes PGx-related knowledge a key component towards precision medicine. A significant part of the state-of-the-art knowledge in PGx is accumulated in scientific publications, where it is hardly reusable by humans or software. Natural language processing techniques have been developed to guide experts who curate this amount of knowledge. But existing works are limited by the absence of a high quality annotated corpus focusing on PGx domain. In particular, this absence restricts the use of supervised machine learning. This article introduces PGxCorpus, a manually annotated corpus, designed to fill this gap and to enable the automatic extraction of PGx relationships from text. It comprises 945 sentences from 911 PubMed abstracts, annotated with PGx entities of interest (mainly gene variations, genes, drugs and phenotypes), and relationships between those. In this article, we present the corpus itself, its construction and a baseline experiment that illustrates how it may be leveraged to synthesize and summarize PGx knowledge., Measurement(s)gene_variant • response to drug • textual entity • chemical entity • haplotype • gene • Pharmacogenomics • Pharmacogenetics • abbreviation textual entity • Pharmacokinetics • Pharmacodynamics • phenotypeTechnology Type(s)digital curationSample Characteristic - OrganismHomo sapiens Machine-accessible metadata file describing the reported data: 10.6084/m9.figshare.11323724

Published
2020
Full Text
View/download PDF

6. Computing Vertex-Vertex Dissimilarities Using Random Trees: Application to Clustering in Graphs

Author

Miguel Couceiro, Kevin Dalleau, Malika Smaïl-Tabbone, Computational Algorithms for Protein Structures and Interactions (CAPSID), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Complex Systems, Artificial Intelligence & Robotics (LORIA - AIS), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), Knowledge representation, reasonning (ORPAILLEUR), Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), ANR-15-RHUS-0004,FIGHT-HF,Combattre l'insuffisance cardiaque(2015), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), and Funded by the RHU FIGHT-HF (ANR-15-RHUS-0004) and the Region Grand Est (France).

Subjects
Theoretical computer science, Computer science, [INFO.INFO-DS]Computer Science [cs]/Data Structures and Algorithms [cs.DS], Random tree, Decision tree, 02 engineering and technology, Complex network, Vertex (geometry), Dissimilarity, Graph clustering, [INFO.INFO-IR]Computer Science [cs]/Information Retrieval [cs.IR], 020204 information systems, Heterogeneous data, 0202 electrical engineering, electronic engineering, information engineering, 020201 artificial intelligence & image processing, Pairwise comparison, [INFO]Computer Science [cs], Cluster analysis, Categorical variable, Attributed graph, Clustering coefficient, MathematicsofComputing_DISCRETEMATHEMATICS
Abstract

International audience; A current challenge in graph clustering is to tackle the issue of complex networks, i.e, graphs with attributed vertices and/or edges. In this paper, we present GraphTrees, a novel method that relies on random decision trees to compute pairwise dissimilarities between vertices in a graph. We show that using different types of trees, it is possible to extend this framework to graphs where the vertices have attributes. While many existing methods that tackle the problem of clustering vertices in an attributed graph are limited to categorical attributes, GraphTrees can handle heterogeneous types of vertex attributes. Moreover, unlike other approaches, the attributes do not need to be preprocessed. We also show that our approach is competitive with well-known methods in the case of non-attributed graphs in terms of quality of clustering, and provides promising results in the case of vertex-attributed graphs. By extending the use of an already well established approach-the random trees-to graphs, our proposed approach opens new research directions, by lever-aging decades of research on this topic.

Published
2019
Full Text
View/download PDF

7. Les forêts d'arbres extrêmement aléatoires : utilisation dans un cadre non supervisé

Author

Kevin Dalleau, Miguel Couceiro, Malika Smaïl-Tabbone, Couceiro, Miguel, Combattre l'insuffisance cardiaque - - FIGHT-HF2015 - ANR-15-RHUS-0004 - RHUS - VALID, Knowledge representation, reasonning (ORPAILLEUR), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), Computational Algorithms for Protein Structures and Interactions (CAPSID), Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Complex Systems, Artificial Intelligence & Robotics (LORIA - AIS), ANR-15-RHUS-0004,FIGHT-HF,Combattre l'insuffisance cardiaque(2015), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), and Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)

Subjects
[INFO.INFO-IR]Computer Science [cs]/Information Retrieval [cs.IR], [INFO]Computer Science [cs], [INFO.INFO-IR] Computer Science [cs]/Information Retrieval [cs.IR], [INFO] Computer Science [cs]
Abstract

National audience; Dans ce travail, nous présentons une nouvelle méthode permettant le calcul de similarités entre objets basée sur les forêts d'arbres extrêmement aléa-toires. L'idée principale de notre méthode est de séparer les données de manière itérative jusqu'à ce qu'une condition d'arrêt soit respectée, et de calculer une similarité basée sur la co-occurrence des instances dans les feuilles de chaque arbre obtenu. Nous évaluons la méthode sur un ensemble de jeux de données synthétiques et réels. Cette évaluation est basée sur la comparaison des similari-tés moyennes entre instances ayant la même étiquette aux similarités moyennes entre instances d'étiquette différente. Ces mesures sont comparables aux notions de similarités intracluster et intercluster, mais ont pour intérêt d'être agnostiques aux choix d'une méthode de clustering en particulier. L'étude empirique montre que la méthode permet effectivement de distinguer les individus n'appartenant pas aux même clusters. Les forêts d'arbres extrêmement aléatoires non supervi-sées ont des propriétés intéressantes, telles que : (i) l'invariance aux transformations monotones de variables, (ii) la robustesse aux variables corrélées, et (iii), la robustesse au bruit. Enfin, nous présentons les résulats obtenus par l'appli-caton d'un algorithme de clustering hiérarchique agglomératif, en utilisant les matrices de similarité obtenues par notre méthode. Les résultats obtenus sur des jeux de données homogènes et hétérogènes sont prometteurs.

Published
2019

8. Unsupervised extremely randomized trees

Author

Malika Smaïl-Tabbone, Miguel Couceiro, Kevin Dalleau, Knowledge representation, reasonning (ORPAILLEUR), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), Computational Algorithms for Protein Structures and Interactions (CAPSID), and Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Complex Systems, Artificial Intelligence & Robotics (LORIA - AIS)

Subjects
Scheme (programming language), Computer science, unsupervised classification, media_common.quotation_subject, Decision tree, 02 engineering and technology, Similarity measure, Machine learning, computer.software_genre, 01 natural sciences, Clustering, 010104 statistics & probability, Empirical research, [INFO.INFO-LG]Computer Science [cs]/Machine Learning [cs.LG], decision tree, 0202 electrical engineering, electronic engineering, information engineering, Quality (business), 0101 mathematics, Cluster analysis, distance, computer.programming_language, media_common, business.industry, ACM: I.: Computing Methodologies/I.5: PATTERN RECOGNITION/I.5.3: Clustering, Running time, similarity measure, ComputingMethodologies_PATTERNRECOGNITION, extremely randomized trees, 020201 artificial intelligence & image processing, Artificial intelligence, business, computer, ACM: I.: Computing Methodologies/I.2: ARTIFICIAL INTELLIGENCE/I.2.6: Learning
Abstract

International audience; In this paper we present a method to compute dissimilarities on unlabeled data, based on extremely randomized trees. This method, Unsupervised Extremely Randomized Trees, is used jointly with a novel randomized labeling scheme we describe here, and that we call AddCl3. Unlike existing methods such as AddCl1 and AddCl2, no synthetic instances are generated, thus avoiding an increase in the size of the dataset. The empirical study of this method shows that Unsupervised Extremely Randomized Trees with AddCl3 provides competitive results regarding the quality of resulting clusterings, while clearly outperforming previous similar methods in terms of running time.

Published
2018

9. Learning from biomedical linked data to suggest valid pharmacogenes

Author

Adrien Coulet, Kevin Dalleau, Patrice Ringot, Sébastien Da Silva, Ndeye Coumba Ndiaye, Yassine Marzougui, Knowledge representation, reasonning (ORPAILLEUR), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL), Ecole Nationale Supérieure des Mines de Nancy (ENSMN), Institut Mines-Télécom [Paris] (IMT)-Université de Lorraine (UL), Interactions Gène-Environnement en Physiopathologie Cardio-Vasculaire (IGE-PCV), Institut National de la Santé et de la Recherche Médicale (INSERM)-Université de Lorraine (UL), ANR PractiKPharma project, grant ANR-15-CE23-0028, funded by the French National Research Agency (http://practikpharma.loria.fr/) and *Snowflake, an Inria associate team (http://snowflake.loria.fr/), Snowflake Inria Associate Team, Inria@SiliconValley, Snowball Inria Associate Team, ANR-15-CE23-0028,PractiKPharma,Confrontation entre connaissances de l'état de l'art et connaissances extraites de dossiers patients en pharmacogénomique(2015), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), Coulet, Adrien, and Interactions humain-machine, objets connectés, contenus numériques, données massives et connaissance - Confrontation entre connaissances de l'état de l'art et connaissances extraites de dossiers patients en pharmacogénomique - - PractiKPharma2015 - ANR-15-CE23-0028 - AAPG2015 - VALID

Subjects
0301 basic medicine, Graph kernel, Computer science, Knowledge discovery from databases, computer.software_genre, [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI], 0302 clinical medicine, False positive paradox, [INFO.INFO-DB] Computer Science [cs]/Databases [cs.DB], [INFO.INFO-BI] Computer Science [cs]/Bioinformatics [q-bio.QM], Valid pharmacogenes, Linked data, Computer Science Applications, Random forest, Identification (information), Phenotype, [SDV.SP.PHARMA] Life Sciences [q-bio]/Pharmaceutical sciences/Pharmacology, 030220 oncology & carcinogenesis, lcsh:R858-859.7, Information Systems, [INFO.INFO-AI] Computer Science [cs]/Artificial Intelligence [cs.AI], Computer Networks and Communications, Health Informatics, [SDV.GEN.GH] Life Sciences [q-bio]/Genetics/Human genetics, lcsh:Computer applications to medicine. Medical informatics, Machine learning, Set (abstract data type), 03 medical and health sciences, [INFO.INFO-LG]Computer Science [cs]/Machine Learning [cs.LG], Computer Graphics, Selection (linguistics), Data mining, Semantic Web, Linkage (software), [INFO.INFO-DB]Computer Science [cs]/Databases [cs.DB], business.industry, Research, Computational Biology, [INFO.INFO-LG] Computer Science [cs]/Machine Learning [cs.LG], 030104 developmental biology, [SDV.GEN.GH]Life Sciences [q-bio]/Genetics/Human genetics, Pharmacogenetics, [SDV.SP.PHARMA]Life Sciences [q-bio]/Pharmaceutical sciences/Pharmacology, Artificial intelligence, [INFO.INFO-BI]Computer Science [cs]/Bioinformatics [q-bio.QM], business, Pharmacogenomics, computer
Abstract

Background A standard task in pharmacogenomics research is identifying genes that may be involved in drug response variability, i.e., pharmacogenes. Because genomic experiments tended to generate many false positives, computational approaches based on the use of background knowledge have been proposed. Until now, only molecular networks or the biomedical literature were used, whereas many other resources are available. Method We propose here to consume a diverse and larger set of resources using linked data related either to genes, drugs or diseases. One of the advantages of linked data is that they are built on a standard framework that facilitates the joint use of various sources, and thus facilitates considering features of various origins. We propose a selection and linkage of data sources relevant to pharmacogenomics, including for example DisGeNET and Clinvar. We use machine learning to identify and prioritize pharmacogenes that are the most probably valid, considering the selected linked data. This identification relies on the classification of gene–drug pairs as either pharmacogenomically associated or not and was experimented with two machine learning methods –random forest and graph kernel–, which results are compared in this article. Results We assembled a set of linked data relative to pharmacogenomics, of 2,610,793 triples, coming from six distinct resources. Learning from these data, random forest enables identifying valid pharmacogenes with a F-measure of 0.73, on a 10 folds cross-validation, whereas graph kernel achieves a F-measure of 0.81. A list of top candidates proposed by both approaches is provided and their obtention is discussed. Electronic supplementary material The online version of this article (doi:10.1186/s13326-017-0125-1) contains supplementary material, which is available to authorized users.

Published
2017
Full Text
View/download PDF

10. Suggesting valid pharmacogenes by mining linked data

Author

Kevin Dalleau, Ndeye Coumba Ndiaye, Adrien Coulet, Knowledge representation, reasonning (ORPAILLEUR), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL), Interactions Gène-Environnement en Physiopathologie Cardio-Vasculaire (IGE-PCV), Institut National de la Santé et de la Recherche Médicale (INSERM)-Université de Lorraine (UL), ANR-15-CE23-0028,PractiKPharma,Confrontation entre connaissances de l'état de l'art et connaissances extraites de dossiers patients en pharmacogénomique(2015), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), Coulet, Adrien, and Interactions humain-machine, objets connectés, contenus numériques, données massives et connaissance - Confrontation entre connaissances de l'état de l'art et connaissances extraites de dossiers patients en pharmacogénomique - - PractiKPharma2015 - ANR-15-CE23-0028 - AAPG2015 - VALID

Subjects
[INFO.INFO-AI] Computer Science [cs]/Artificial Intelligence [cs.AI], pharmacogenomics, [INFO.INFO-DB]Computer Science [cs]/Databases [cs.DB], Link prediction, [INFO.INFO-LG] Computer Science [cs]/Machine Learning [cs.LG], [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI], [INFO.INFO-LG]Computer Science [cs]/Machine Learning [cs.LG], Linked Data, pharmacogenes, [INFO.INFO-DB] Computer Science [cs]/Databases [cs.DB], [INFO.INFO-BI]Computer Science [cs]/Bioinformatics [q-bio.QM], Data mining, Linked Open Data, random forest, [INFO.INFO-BI] Computer Science [cs]/Bioinformatics [q-bio.QM]
Abstract

International audience; A standard task in pharmacogenomics research is identifying genes that may be involved in drug response variability, i.e., pharmacogenes. Because genomic experiments tended to generate many false positives, computational approaches based on the use of background knowledge have been proposed. Until now, those have used only molecular networks or the biomedical literature. Here we propose a novel method that consumes an eclectic set of linked data sources to help validating uncertain drug–gene relationships. One of the advantages relies on that linked data are implemented in a standard framework that facilitates the joint use of various sources, making easy the consideration of features of various origins. Consequently, we propose an initial selection of linked data sources relevant to pharmacogenomics. We formatted these data to train a random forest algorithm , producing a model that enables classifying drug–gene pairs as related or not, thus confirming the validity of candidate pharmacogenes. Our model achieve the performance of F-measure=0.92, on a 100 folds cross-validation. A list of top candidates is provided and their obtention is discussed.

Published
2015

Catalog

Books, media, physical & digital resources
See catalog results